What population do you think I was sampling from, exactly? And your captiousness is not contributing to the discussion, which was to point out that the subscription/profit variable is only weakly correlated with being shut down, contrary to claims that free services are uniquely vulnerable.
